The short answer
The Marcus by Goldman Sachs file has 6 columns, dates written as YYYYMMDD and one amount column where a positive value is money out. The CSV written for Wave has the columns Date,Description,Amount. So the date is rewritten as MM/DD/YYYY, the amount is written as one signed number, and the text columns are moved as in the table below.
Paid profile. The free manual column mapper works for any bank, this one included, but the free version writes only the first 25 converted rows of a file. The 9 USD one-time licence adds the ready-made Marcus by Goldman Sachs profile (one of the 35 paid profiles) and removes the 25-row limit. The column table on this page is free to read and can be used to fill in the manual mapper.
The converter runs in your browser. Your file is not uploaded.
What the Marcus by Goldman Sachs CSV looks like
- Country: United Kingdom.
- Account type: savings (Marcus UK offers savings accounts; the source does not name the product).
- Header row: yes, one row before the first transaction.
- Columns: 6, separated by commas.
- Date format: YYYYMMDD, for example 20240304 (the source writes it as Ymd).
- Amounts: one amount column where a positive value is money out. Evidence for the direction: it is inferred from the role the open-source importer gives the column; no sample row confirms it.
- Header text: the source gives column positions and meanings only, not the header names.
Note kept with this layout in the profile data: Dates have no separators (20240305). The sign comes only from the importer treating the amount column as a debit; for a savings account this is uncertain, so the converter should ask the user to confirm the direction in the preview.
What Wave accepts
- File type: CSV.
- Header row: not stated in the excerpts read; the converter writes one.
- Required columns: Date, Description, Amount.
- Amounts: Three or four columns: date, description, and the amount either all in one column or one column each for deposits and withdrawals. For credit card statements, purchases must be positive and payments negative.
- Dates: MM/DD/YYYY.
| Wave column | Required by Wave | Filled from the Marcus by Goldman Sachs file |
|---|---|---|
| Date | required | column 1 |
| Description | required | column 2 |
| Amount | required | column 3 |
From a web search excerpt of the vendor's help page (the page itself refuses automated reading; re-check it in a browser): "To import properly, your CSV file must have three or four columns: date, amount of the transaction (all in one column, or one column each for deposits and withdrawals), and description."
Column-by-column mapping
Every column of the Marcus by Goldman Sachs file, left to right, and where it ends up in the Wave file.
| Column | Header text | What it holds | Where it goes in the Wave file |
|---|---|---|---|
| 1 | not given by the source | Transaction date | Date, rewritten as MM/DD/YYYY |
| 2 | not given by the source | Description | Description |
| 3 | not given by the source | Amount (a positive value is money out) | Amount, with the sign turned so that money out is negative |
| 4 | not given by the source | Not used | Not copied |
| 5 | not given by the source | Not used | Not copied |
| 6 | not given by the source | Account or card identifier | Not copied |
Pitfalls for Marcus by Goldman Sachs to Wave
Dates
- The date must be reordered. The bank file is written year first, then month, then day (YYYYMMDD, for example 20240304). The Wave file is written month first, then day, then year (MM/DD/YYYY, for example 03/04/2024).
- The date has no separators (20240304). A spreadsheet opens it as a plain number, so it must be split into year, month and day by position.
- Wave on dates: MM/DD/YYYY. As stated by the vendor: the date column must be in MM/DD/YYYY format.
Amounts and their direction
- Sign flip: in this bank file a positive amount is money out, and the Wave file needs money out as a negative number. Every amount is multiplied by -1.
- Wave's own rule: Three or four columns: date, description, and the amount either all in one column or one column each for deposits and withdrawals. For credit card statements, purchases must be positive and payments negative.
- How sure the direction is: it is inferred from the role the open-source importer gives the column; no sample row confirms it. Compare two rows you recognise with the preview before importing; the converter has a switch that flips the sign of every amount.
Other Wave import rules
- Keep only the date, amount and description columns; delete any other columns.
- The description column can be empty, but if filled it must contain words, not only numbers.
- Remove special characters such as #, &, $ and * from the file.
- Files with more than 100 rows tend not to upload; split them into files of no more than 100 rows each.
Before and after: an invented example
Invented sample data. These four rows were made up for this page. They are not from a real Marcus by Goldman Sachs export and hold no real account, name or balance.
The source gives column positions only, so the headings below are the column number and what the column holds, not the bank's header text. The 2 unused columns of the file are left out of the table.
| 1: transaction date | 2: description | 3: amount | 6: account or card identifier |
|---|---|---|---|
| 20240304 | COFFEE CORNER MAIN ST | 4.50 | 0000 |
| 20240305 | PAYROLL EXAMPLE CO | -2500.00 | 0000 |
| 20240307 | RENT MARCH | 1234.56 | 0000 |
| 20240311 | REFUND EXAMPLE SHOP | -15.25 | 0000 |
| Date | Description | Amount |
|---|---|---|
| 03/04/2024 | COFFEE CORNER MAIN ST | -4.50 |
| 03/05/2024 | PAYROLL EXAMPLE CO | 2500.00 |
| 03/07/2024 | RENT MARCH | -1234.56 |
| 03/11/2024 | REFUND EXAMPLE SHOP | 15.25 |
Do it by hand in a spreadsheet, free
These steps need no tool from this site. The cell references follow the column positions of the profile above; if your spreadsheet uses semicolons between formula arguments, replace the commas.
- Open the Marcus by Goldman Sachs CSV in a spreadsheet program through its import dialog. Set the date column (column 1) to Text so the program does not reinterpret it, and leave the amount columns as numbers.
- Row 1 is the header and the transactions start in row 2.
- Date: rewrite column A from YYYYMMDD to MM/DD/YYYY (20240304 becomes 03/04/2024). In a new column, for the first transaction:
=MID(A2,5,2)&"/"&MID(A2,7,2)&"/"&MID(A2,1,4)This formula reads the date by character position, so it only works while the dates are text and every date has a two-digit day and month. Fill it down. - Amounts: in this file a positive value is money out, so turn every sign:
=-C2 - Text columns: fill Description (from column 2). Where several bank columns feed one column, join them with a separator such as " - ".
- Apply Wave's own rules from its help page: Keep only the date, amount and description columns; delete any other columns. The description column can be empty, but if filled it must contain words, not only numbers. Remove special characters such as #, &, $ and * from the file. Files with more than 100 rows tend not to upload; split them into files of no more than 100 rows each.
- Keep only the new columns, put the header row Date,Description,Amount on top, save the sheet as a comma-separated CSV, and compare the dates, the direction of the amounts and the number of rows with your statement in Wave's import preview.
Or use the converter
Load the file, choose the Marcus by Goldman Sachs profile in the bank list and the target Wave, read the preview, then download. The conversion runs locally in your browser: the file is not uploaded, and only the licence key is sent for unlocking.
Paid profile. The free manual column mapper works for any bank, this one included, but the free version writes only the first 25 converted rows of a file. The 9 USD one-time licence adds the ready-made Marcus by Goldman Sachs profile (one of the 35 paid profiles) and removes the 25-row limit. The column table on this page is free to read and can be used to fill in the manual mapper.
Where this layout comes from
No bank supplied or approved this layout. It was copied on 5 October 2026 from a public open-source importer file kept by volunteers: firefly-iii/import-configurations, file gb/marcus/default.json
The lines used from the source: "headers": true, "date": "Ymd", roles: date_transaction, description, amount_debit, _ignore, _ignore, account-number
The Wave rules come from the vendor's own help pages: support.waveapps.com/hc/en-us/articles/208621626-Upload-a-bank-or-credit-card-statement-in-csv-format (named as text, not linked), support.waveapps.com/hc/en-us/articles/208622596 (named as text, not linked). The official page refuses automated fetches (HTTP 403 or a script-only page), so the rules were read from web search excerpts of that same official URL on the checked date. Re-check in a browser before relying on a detail.
Bank formats change. Banks alter their CSV exports without notice and vendors change their import rules, so this mapping can be out of date. If the preview looks wrong, use the manual column mapper.
Related pages
Marcus by Goldman Sachs to other apps
The Marcus by Goldman Sachs guides for YNAB, QuickBooks Online and Xero, in the bank list